FUNDAMENTALS OF MARKET RISK

The following is a very brief summary of some of the specific risks associated with each of the main
categories of Market Risk.

EQUITY RISK

• The volatility of prices of traded equities – according to the data from MSCI (see Figure 1)
the annualized standard deviation for the MSCI World Index over the last 10 years is almost
13% per annum and for Emerging Markets around 18%
• The maximum drawdown (discussed elsewhere in this module) for the MSCI World Index
occurred between the peak in October 2007 and the trough in March 2009. To express this
simply if one had invested in the index at the peak and liquidated at the trough then a loss of
almost 58% would have been realized. In the case of the EM index this would have been
65%
• Equity investors are exposed to capital losses which can arise from bankruptcy,
reorganizations and market crashes
• Income losses for equity investors can arise from dividend suspension

Figure 1 Risk and reward characteristics of the major MSCI equity indices

INTEREST RATE RISK

• Figure 2 highlights how US three-month T-bill rates – steered by the Federal Reserve as part
of its monetary policy – have fluctuated over the last few decades
• Re-pricing and re-investment risks arise due to fluctuations in short term/long term rates
• Changes in the shape of the yield curve pose specific risks for fixed income managers and
can also be harbingers of broader macro-economic developments.

Information Classification: General

 Figure 2 The US Treasury Bill rate since 1955

FOREIGN EXCHANGE RISK

• Figure 3 shows the EUR/USD exchange rate since the currency’s inception in 1999 and the
rate has moved between below parity in the earlier period and reached a peak close to
$1.60 in late 2007
• Banks must address, on a daily basis, the risks within the Trading Book of adverse currency
movements in assets denominated in currencies other than the domestic or base currency
• Minor currencies and EM currencies can be quite volatile and from time to time illiquid

Figure 3 EUR/USD exchange rate since Inception

COMMODITY RISK

• Many commodities exhibit strong volatility. Figure 4 shows the price of crude oil (West
Texas Intermediate) over an extended period. In late 2007 the price reached almost $150
per barrel whereas in the late summer of 2008 this had dropped to just over $30 per barrel
• In commodity futures markets it is not uncommon for prices to move substantially from day
to day and even intraday. Limit down (or up) moves can occur which pose liquidity and
execution risks for traders in such markets
• Margin requirements for futures trades can change quickly depending on the volatility of the
underlying commodity.

Information Classification: General

 Figure 4 Crude Oil prices: West Texas Intermediate

WHAT IS THE MARKET RISK PREMIUM?

To answer this question – fundamental to financial risk theory – we need to touch on the Capital
Asset Pricing Model (CAPM). This model was developed by several American economists during the
1970’s including William Sharpe and Fischer Black. The model is still considered to be the
foundation for explaining the expected return on financial assets and relating risk to reward.

The CAPM proposes that to understand the reward that investing in marketable securities provides –
the market risk premium – we need to create a Securities Market Line based on two axes – the
vertical axis represents the return on securities or a portfolio and the horizontal axis represents the
level of risk (known as beta). The diagram below, Figure 5, provides an intuition into how the
market risk premium is determined.

Figure 5 Market Risk Premium

The risk-free rate (RFR) is the returned for a risk averse investor – it is equivalent to the return
earned on short term government securities where there is no default risk. Beta (as will be
explained below) is a measure of risk, and where the return earned by an investor who has exposure
to the overall market is equivalent to the beta value of one. For the risk averse investor earning the
RFR the beta value is zero. The Market Risk premium is the excess return earned above the risk-
free rate by taking on the risk of owning marketable securities

Information Classification: General

 BETA AND SYSTEMATIC RISK

Figure 6 Deconstructing Market Risk

Figure 6 illustrates one of the foundation ideas of the CAPM and establishes the important notion
that financial risk can be deconstructed into two distinct kinds of risk. On the one hand there is
systematic or market risk which is the risk that all holders of financial assets have by deciding to own
a financial claim for which there is a fluctuating price. Within the CAPM, holding market securities
that are subject to risk is contrasted with holding risk free securities such as US Treasury bills. In
return for being willing to incur the risk of securities other than T bills market participants are
rewarded with a market risk premium. It is at their own discretion of how much of a risk premium
they will earn based upon how much risk they are prepared to tolerate. In essence the systematic or
market risk from holding a collection of securities cannot be removed by diversification of the asset
mix, and, in order to manage systematic risk we have to engage in some form of hedging exercise,
which will be very much the focus of another module in this course.

On the other hand there is unsystematic or idiosyncratic risk and this risk can be attributed to the
variability in the returns of specific individual securities. Individual securities will have their own
idiosyncratic risks – in the case of equities, companies may report poor earnings and in a worst case
go bankrupt. One of the key ideas of Modern Portfolio Theory was that idiosyncratic risk can be
substantially reduced by creating a diversified portfolio. Markowitz demonstrated the benefits of
combining assets within a portfolio and provided a robust conceptual framework relating the co-
variances (or correlations) of the portfolio’s constituents to the overall variance or volatility of the
combination.

A key insight that originated with Markowitz and then became a cornerstone concept in the CAPM
was the notion that financial risk could be deconstructed into two distinct kinds of risk. On the one
hand there is the risk that all holders of financial assets have simply by deciding to own a financial
claim for which there is a fluctuating price. This type of risk is known under the CAPM as the
systematic risk or market risk. In contrast there is unsystematic or idiosyncratic risk which is
attributable to the variability in returns of a specific individual security.

Information Classification: General

 Figure 7 Systematic and Unsystematic Risk – the CAPM under normal market conditions

The first thing to be noted is that Figure 7 is applicable under normal market conditions. When
financial markets are behaving abnormally the diagram takes on different characteristics stemming
from the disappearance of diversification when markets become highly correlated.

The horizontal axis of Figure 8 shows the number of individual securities that are held within a
portfolio – it could be individual equities or a mixture of assets. For simplicity let us assume that we
are considering the riskiness of an equity portfolio and that the securities are in fact holdings of
ordinary shares for a variety of companies.

The blue line represents the systematic risk (i.e. the volatility of returns) for equities (say UK
equities) and it has been shown at 10%. The red line corresponds to the standard deviation of
returns resulting from the number of holdings shown on the horizontal axis – this is the
unsystematic risk of the portfolio.

As the number of holdings approaches twenty the CAPM theory maintains that the unsystematic risk
will have more or less converged with the systematic risk. As illustrated in the diagram the red line
becomes asymptotic with respect to the blue line and the reduction in unsystematic risk by
addingfurther securities beyond the threshold level of 20 becomes vanishingly small.

On the right hand scale (RHS) we can read off the percentage for returns (green line values) which
correspond to the level of holdings. Returns will fluctuate but will eventually – beyond 20 securities
– converge to the overall market mean return.

Information Classification: General

 The key idea is that from the perspective of an investor holding assets there are two types of market
risk which, in accordance with the CAPM, can be distinguished. In focus at this point are metrics
relating to the mean returns and the variance, or more typically the standard deviation of returns, to
an overall market such as equities or (say) UK gilts or US Treasury bonds. The variance represents
the degree to which the set of individual returns over a specified period differ or deviate from the
average return. The variance captures the scatter or dispersion of such returns. In a sense it is a
measure of the waywardness of returns or the likelihood that one will encounter incidences where a
rogue return will occur.

The assumption implicit in this notion of risk is that the investor will be averse to receiving negative
return surprises and would therefore be inclined to seek out those assets where the dispersion from
the mean returns are tamer and less likely to produce negative surprises. Of course, in so doing one
may almost certainly be reducing the likelihood of positive surprises as in a distribution of returns
with relatively low skew the prevalence of returns which are beneficial i.e. will produce generous
returns is roughly symmetrical with the prevalence of loss-making returns.

Also, worth pointing out is that there is no regard to the question about the expected holding period
for the investor. For the patient long term investor it could be argued that there is far less concern
about the variability of returns rather it is the underlying long term trend of the mean returns which
is critical. For an investor or trader with a shorter time frame the “risk” is higher in that there is
increased likelihood that the investor would be realizing a loss if the variability of returns is relatively
high and the liquidation of assets had to occur at an inopportune moment.

The best way to illustrate the distinction is by contrasting the “risk” that would be confronted by a
passive investor that owns an index fund – such as a FTSE 100 or S&P 500 tracker – and the active
investor that has created a portfolio with securities selected from all available securities in a market.
The passive investor is only subject to systematic risk whereas the active investor takes on both the
systematic risk of holding assets in general as well as the additional unsystematic risk based on the
idiosyncratic risks associated with the set of securities that have been selected for the portfolio.

Diversification becomes one of the primary methods for risk management. If a portfolio contains a
variety of securities – for many of which there is relatively weak correlation in the trajectories of
returns – the benefits of diversification are that while some securities may be suffering from adverse
price developments and volatility there will be other securities which are behaving in a more
advantageous fashion for the investor. The amalgamation of such a variety will ameliorate the
outliers of the less well-behaved instruments.

When returns become too highly correlated there is a corresponding drop in the benign
consequences attributable to diversification. If most securities are following the same returns path,
then there is not much to choose between owning a few securities or many…they will all be rising or
falling in unison.

The main factor which will drive the diversification benefits of a portfolio and which will, in turn,
determine the volatility of that portfolio, or its standard deviation, is the extent to which the returns
of the assets combined are more or less correlated. Since a good part of our discussion elsewhere in
this course will be on the pivotal role that both correlations amongst asset returns and volatility
have in determining Value At Risk and also hedging, it is worth examining in a relatively simple
manner the process of combining assets and the key statistical values resulting from such
combinations.

Information Classification: General

 COMBINING ASSETS AND CALCULATING THE PORTFOLIO STANDARD DEVIATION

The key formula for calculating the standard deviation for a two asset portfolio is shown below. The
logic can be seen in the Excel workbook Expected Returns.xlsx.

𝑉𝑎𝑟𝑖𝑎𝑛𝑐𝑒 𝑜𝑓 𝑃𝑜𝑟𝑡𝑓𝑜𝑙𝑖𝑜 = 𝑝𝑎2 ∗ 𝜎𝑎2 + 𝑝𝑏 2 ∗ 𝜎𝑏 2 + 2 ∗ 𝑝𝑎 ∗ 𝑝𝑏 ∗ 𝜎𝑎 ∗ 𝜎𝑏 ∗ 𝐶𝑜𝑟(𝑎𝑏)

Where pa = the proportion of the portfolio allocated to security A

pb = the proportion of the portfolio allocated to security B

σ² = the variance of the security

Cor = the coefficient of correlation between the expected returns of A and B

σ = the standard deviation of the security

The above formula can be applied to two Assets A and B which, let us assume arbitrarily, have a
correlation coefficient of 0.49. The standard deviations of monthly returns for each asset are also
calculated within Excel and shown in the table below.

By combining assets A and B into a portfolio with a ratio of 50:50 the standard deviation of the
portfolio has been reduced to a value below that for holding either 100% of asset A or 100% of asset
B. The actual monetary returns for each of A, B and A&B are also shown in the table below, with the
compounded monthly returns being given as well.

Figure 8 Positively correlated returns

From Figure 8 a 50/50 combination of asset A and asset B provides an annualized compounded
return of 6% and an annualized standard deviation of 10.28%. It can also be seen from the table that
a portfolio which only contains asset A would show an annualized compounded return of 5.7% but
with a higher standard deviation in the returns of 11.42%, whereas a portfolio which consists only of
asset B would have an annualized compounded return of 5.92% and an annualized standard
deviation of 12.36%.

Information Classification: General

 Figure 9 Negatively correlated returns

By combining negatively correlated assets, in the manner illustrated in Figure 9, it is possible to

achieve a significant reduction in the overall risk of the portfolio, while still achieving a relatively high
expected return. The standard deviation of the returns for the 50/50 portfolio of A and B shown in
Figure 9 is lower than the standard deviation for holding just A or just B and yet the A&B portfolio
should generate more than an 8% return which is considerably above the roughly 2% return from
holding just B.

BETA IS NOT THE SAME AS VOLATILITY

As mentioned in our review of the CAPM, beta or β is a measure of the systematic risk of a portfolio
compared to the risk of the market portfolio. It indicates the risk sensitivity of an asset/portfolio
with respect to the market as a whole or a suitable benchmark index:

where 𝜎P is the sigma or standard deviation of the portfolio

𝜎M is the sigma or standard deviation of the overall market or benchmark index

Correl(P, M) represents the correlation coefficient between the portfolio return and the market
returns, and β is beta

If we assume, and until recently this was not far from the case, that the risk-free rate is zero and that
an equity portfolio has a beta value of 1.2 this indicates that, if the benchmark returns +10% then
the portfolio will return +12%, and likewise if the benchmark has a negative return e.g. -10%, the
portfolio will return -12%. A beta value in excess of one indicates that the systematic risk of the
portfolio i.e. the non-diversifiable characteristics of the portfolio, makes it “riskier” than the overall
market. It highlights the intrinsic nature of financial risk/reward which is that to be eligible for a
higher return one must be ready to accept more risk.

Information Classification: General

 Beta is a method of scaling the risk premium and it can thus be stated that the expected return for a
portfolio is equal to the risk-free rate plus a multiple of the equity risk premium

Re = Rf + (Rm - Rf)
Re = expected or required return of shareholders

Rf = risk-free rate

Rm = return on the market portfolio

 = measure of systematic risk relative to the market average

Beta can be found for an asset by determining the slope of the linear regression where the returns
from a portfolio (the dependent variable) are examined against the returns from a selected
benchmark (the independent variable).

Beta becomes a critical variable in determining how to manage the risk of holding a portfolio of
assets and how to hedge that portfolio from systematic risk. To put things simply if we are holding a
portfolio which is equivalent to the whole market then we are subject to the volatility of the market
and should earn the market risk premium. If we hold a portfolio which has a beta of more than 1
then we should expect a higher return (or loss) than the market risk premium.

Information Classification: General

 BETA ADJUSTMENT AS A MEASURE OF RISK MANAGEMENT
The beta of an asset is in many ways one of the most useful variables from the perspective of
managing the risk of holding financial assets. There is a certain corollary with the concept of
duration which was explored in Module 2. Making an adjustment in the duration of a fixed income
portfolio based upon the outlook for interest rates is not only a useful strategy from the point of
view of asset allocation, but also can insulate a portfolio from adverse price movements and
drawdowns.

With respect to an asset’s beta it can be seen that if one has a bullish outlook on equity prices –
other things being equal, i.e. the idiosyncratic qualities of the securities selected are not especially
problematic – then having a relatively high beta portfolio will enable an asset manager to
outperform the benchmark.

If the outlook for equities is bearish, one would want to adjust the beta downwards or, if one is
aggressive and the investment mandate permits it, one would possibly want to include some short
positions in high beta equities within the overall portfolio mix. Of course, the efficacy of beta
adjustment, from both an asset allocation and risk management point of view, will depend on
whether the outlook in fact is correct and equities perform in accordance with one’s forecast

HEDGE RATIO

One method which enables us to hedge an equity portfolio is to use stock index futures contracts
which will be discussed in detail elsewhere.

A futures hedge ratio indicates the number of contracts needed to emulate the behavior of a
portfolio. The hedge ratio has two components:

– The scale factor which addresses the monetary value of the portfolio relative to the
monetary value of the futures contract
– The level of systematic risk i.e., the beta of the portfolio
–
𝑴𝒐𝒏𝒆𝒕𝒂𝒓𝒚 𝑽𝒂𝒍𝒖𝒆 𝒐𝒇 (𝑳𝒐𝒏𝒈)𝑷𝒐𝒓𝒕𝒇𝒐𝒍𝒊𝒐
𝑯𝒆𝒅𝒈𝒆 𝑹𝒂𝒕𝒊𝒐 (𝑯𝑹) = 𝒙 𝑩𝒆𝒕𝒂
𝑴𝒐𝒏𝒆𝒕𝒂𝒓𝒚 𝑽𝒂𝒍𝒖𝒆 𝒐𝒇 𝑩𝒆𝒏𝒄𝒉𝒎𝒂𝒓𝒌

Let us look at a specific example for hedging an equity portfolio where we have calculated the beta
of the portfolio with reference to the S&P 500 index. (It is worth pointing out that many of the
major global equity indices such as the DAX, CAC 40, and FTSE 100 are very highly correlated to the
performance of the S&P 500 so it is not US centric to use futues contracts based on the S&P 500
especially as it is the most liquid of the global equity index futures.)

The following then are the assumptions:

• Value of equities portfolio to be hedged = $70 million

• Estimated portfolio beta with respect to the S&P 500 index = 1.25
• Value of one S&P 500 contract = $250 x S&P 500 Cash price
• Current S&P 500 price = 2800 therefore single contract = $700,000

The number of contracts to sell of S&P 500 futures contracts

= {$70,000,000/$700,000} * 1.25 = 125 contracts

10

Information Classification: General

 In this scenario we would be hedging the cash portfolio provided that the beta value we calculated
remains stable during the period of the hedge. In fact, as will be discussed later, beta values are not
stable and this introduces the need for periodic re-balancing of the hedge position by re-
calculating the hedge ratio from time to time. Re-balancing becomes even more of an issue when
considering options as a hedging instrument as they have further issues related to the sensitivity of
option prices to changes in the underlying instrument (i.e. the delta of an option) as well as time
decay of option contracts.

INTRADAY MARKET RISK

Let us turn now to a kind of market risk which is often not addressed in the risk management
literature but which poses considerable challenges for those running a trading book and especially
where intraday risk management protection protocols may be violated.

The discussion revolves around an episode known as the Flash Crash and it occurred on 6 May 2010,
and because of its severity it became the subject of a detailed forensic analysis by the US regulators
– the SEC and CFTC who produced a very substantial report on the event. It is also worth pointing
out that similar episodes have occurred since 2010 although not as severe. Some market
commentators believe that the risk of future episodes remains reasonably high.

• On 6 May 2010, disruption to trading in the S&P 500 futures contracts (mainly in the Globex™
traded e-mini contract) caused the S&P 500 to drop by almost 100 points and the Dow Jones
Industrial Average dropped by almost 1,000 points. In both cases, this was a drop of between 8-
9%.

• The so called “Flash Crash” lasted for about 60 minutes and started around 14:30 EST.

• Before the end of that day of trading there was a sizable recovery – the S&P 500 finished the
session down only 3% for the day.

• Many exchange traded funds (ETF’s) showed massive downward spikes.

• The initial blame for the crash was attributed to a “fat finger” trade – subsequently this was
totally discredited.

• The SEC/CFTC inquiry provided a far more detailed investigation, but in the opinion of many it still
failed to identify the trigger which caused the crash.

• The event turned out not to have been an aberration as there was a significant subsequent
correction in global equity indices during the early summer of 2010.

• Although 6 May 2010 was the third-highest-volume-day in the history of e-mini S&P 500 futures,
there is consensus in characterizing it as an extremely illiquid day.

• The SEC/CFTC report stressed that “high trading volume is not necessarily a reliable indicator of
market liquidity”.

Figure 11 below, which comes from the SEC/CTFC report on the crash, shows how dramatically the
futures contracts fell and also the massive spike in volume that coincided with the drop. The data

11

Information Classification: General

 points on the horizontal axis are for 60 seconds and, as can be seen, there was a vicious downward
spiral where selling begat more selling. The SEC/CFTC report suggested that the disruption within
the equity markets – and in such an inter-connected financial system, it spilled over into other
markets – was exacerbated by the nature of high frequency trading (HFT) algorithms.

In Module 4 on liquidity risk, we shall consider HFT algorithms in more depth, but at this point it will
suffice to say that electronic intermediaries using very fast trading algorithms have replaced almost
all of the traditional market making function within financial markets. In its traditional sense market
making was the provision of liquidity within markets by human agents/brokers who stood ready to
buy and sell in order to facilitate trade executions with adequate liquidity. What appears to have
happened on 6 May 2010 was that a disruptive event – perhaps triggered by a large move in the FX
market – created a positive feedback loop in which electronic liquidity providers soon became
liquidity consumers as prices continued to spiral downward.

Figure 11

When all market participants wanted to be on the sell side and there were no or few bids then prices
had to collapse. In subsequent analysis it appears that more “normal” market conditions were only
resumed when the HFT algorithms were “turned off” or suspended.

The reason for discussing this phenomenon in some detail is that it clearly has serious implications
for intraday risk management on a trading desk. Most trading desks operate with stop loss levels
where, if a certain loss threshold is breached, then the position needs to be closed out.

12

Information Classification: General

 As a consequence of the Flash Crash, many companies’ shares dropped precipitously as did a large
number of very liquid exchange traded funds (ETF’s) and automatic stop loss triggers could have
produced substantial losses. In fact, in the aftermath of the crash, the trading exchanges decided to
cancel many of the trades which occurred at the more extreme price levels that were reached during
that day of trading.

The question for risk managers is whether a risk control protocol requiring such non-discretionary
stop losses is prudent or whether they should be held in abeyance until the end of a trading day. It
is not necessarily as easy to answer this as it may seem as it is also not easy to determine whether a
serious disruption to market prices is just a temporary phenomenon or the beginning of a more
extended crisis.

MODERN PORTFOLIO THEORY (MPT) AND DIVERSIFICATION

Modern portfolio theory (MPT) originated with an article by Harry Markowitz entitled Portfolio
Selection that was published in the Journal of Finance in 1952. The principles that were proposed in
this highly influential paper now seem to be unremarkable, but at the time the emphasis on
diversification and the mathematical model that enabled a portfolio manager to assemble positions
which will provide the best level of expected return for a given risk tolerance, helped to usher in a
period of great innovation in investment theory. The MPT foundations were a cornerstone of the
work of William Sharpe that culminated in the CAPM. Markowitz, Sharpe and a further collaborator
Merton Miller shared the Nobel Prize for Economics in 1980.

Markowitz’s seminal insight and contribution to asset allocation theory and risk management was
his provision of a quantitative technique that encouraged the astute portfolio manager to focus on
selecting portfolios based on their overall risk-reward characteristics rather than constructing
portfolios from consideration of only their individual profit opportunities.

MPT stresses that an investor should allocate assets based on the characteristics of a portfolio rather
than the individual characteristics of the constituent securities considered separately. Prior to the
MPT, the received wisdom on the manner to combine securities in a portfolio was to screen
securities that offered the most attractive opportunities for gain with the least risk and then add
these together in a portfolio. Bringing individual securities together in such a fashion would often
lead to exposing the portfolio to too many securities from the same sector where the correlations
between the returns amongst the securities selected would be imprudently high. In other words, the
portfolio would lack the benefits of diversification. Markowitz’s major contribution was to articulate
the logic of diversification and to focus attention on how the overall volatility of a portfolio (which is
a suitable proxy for its degree of risk) is calculated from the covariance matrix of the returns of its
constituents. A discussion of the logic behind these calculations is covered elsewhere in this course.

After Markowitz’s paper an asset manager had a systematic procedure for evaluating different
combinations of securities and selecting those combinations which provided the optimal reward for
a given level of risk. The optimal allocations will be a trade-off between the risks one is willing to
tolerate and the anticipated returns. The Markowitz procedure enables a fund manager to calculate
from the observed (expected) correlations of its constituents, the overall portfolio volatility and the
expected returns for numerous combination scenarios. From the logical space of possible portfolio
combinations there are a series of combinations that will optimally balance the risk and reward. The
optimal combinations that maximize the reward for the different possible levels of risk lie on what
Markowitz termed the efficient frontier. The fund manager can settle on the level of risk that is
acceptable and then select the specific combination of securities that optimize the reward for this
level of risk.

13

Information Classification: General

 As can be seen in the graphic in the top left part of the panel in Figure 12 the efficient frontier
captures those allocations between A and B which produce a superior reward for a given level of
risk. At any point below the efficient frontier – for example if the portfolio consisted of only A then
the risk (i.e. the standard deviation) is 5% but only generates a 10% return. If however we increase
the allocation to B - somewhere around 45% to B (see the row which corresponds to a weighting of
A around 55%) - the portfolio has a risk of 5% but now generates a return of around 15%. Moving
further up the efficient frontier from this point shows that greater risk is being rewarded with a
higher return.

Figure 12 The Efficient Frontier

14

Information Classification: General